EMC® VPLEX™ Host Multipathing

Implementation Planning and Best Practices Technical Notes
EMC® VPLEX™ Host Multipathing

Implementation Planning and Best Practices
 ESXi NMP Policy
 PowerPath
 Native MPIO
 Updated for GeoSynchrony 6.0 and VS6 Hardware Platform
Abstract
This document describes multipath options and settings for use with VPLEX in Local,
Metro and host cross-cluster configurations.
August 2016
Copyright © 2016 EMC Corporation. All rights reserved. Published in the USA.
Published September 2016
EMC believes the information in this publication is accurate as of its publication date.
The information is subject to change without notice.
The information in this publication is provided as is. EMC Corporation makes no

representations or warranties of any kind with respect to the information in this
publication, and specifically disclaims implied warranties of merchantability or
fitness for a particular purpose. Use, copying, and distribution of any EMC software
described in this publication requires an applicable software license.
EMC2, EMC, and the EMC logo are registered trademarks or trademarks of EMC
Corporation in the United States and other countries. All other trademarks used
herein are the property of their respective owners.
For the most up-to-date listing of EMC product names, see EMC Corporation
Trademarks on EMC.com.
EMC VPLEX Host Multipathing

Part Number H13547
2 EMC VPLEX Host Multipathing

Contents
Contents
Chapter 1 Introduction 5
Audience .................................................................................................................... 6
Document Organization .............................................................................................. 6
Host Multipath for VPLEX Overview ............................................................................. 7
Multipath Software ................................................................................................ 8
Chapter 2 ESXi NMP Policy 10

VMWARE ESXi ........................................................................................................... 11
Round Robin - New Policy Recommendation ............................................................. 11
Round Robin ........................................................................................................ 11
Changing Policy from Fixed Path to Round Robin ...................................................... 12
Changing Policy ................................................................................................... 12
Additional Information ............................................................................................. 12
Chapter 3 PowerPath 14
Introduction ............................................................................................................. 15
Path management ............................................................................................... 15
PowerPath Features ............................................................................................. 15
Supported Versions and Settings ............................................................................. 16
Chapter 4 Native MPIO 17

Introduction ............................................................................................................. 18
IBM AIX..................................................................................................................... 19
HP-UX ....................................................................................................................... 21
Linux ........................................................................................................................ 22
Red Hat Enterprise Linux (RHEL) ........................................................................... 22
SuSE Linux Enterprise Server (SLES) .................................................................... 27
Microsoft Windows ................................................................................................... 31
Windows 2008 Server Core and Windows 2008 R2 .............................................. 31
Server Core .......................................................................................................... 31
Path management in Multipath I/O for VPLEX ...................................................... 32
Enabling Native MPIO .......................................................................................... 32
Native MPIO with Windows Server 2012 .............................................................. 33
Configuring Native MPIO for Windows Server 2012 .............................................. 33
Device discovery and claiming devices for MPIO .................................................. 34
EMC VPLEX Host Multipathing 3

Contents
Path management in Multipath I/O for VPLEX ...................................................... 35

Solaris ...................................................................................................................... 37
VPLEX and MultiPath software ............................................................................. 37
VPLEX and MPxIO ................................................................................................. 39
VPLEX with Oracle Solaris Cluster support ............................................................ 40

Chapter 1: Introduction
Chapter 1 Introduction
This chapter presents the following topics:
Audience ................................................................................................................. 6
Document Organization ........................................................................................... 6
Host Multipath for VPLEX Overview .......................................................................... 7

Audience
These technical notes are for EMC field personnel and partners and customers who
will be configuring, installing, and supporting VPLEX. An understanding of these
technical notes requires an understanding of the following:
 SAN technology and network design

 Fiber Channel block storage concepts
 VPLEX concepts and components
Document Organization
This technical note is one of a set of documents that supersede the monolithic
Implementation Planning and Best Practices for EMC VPLEX Technical Notes
document that had previously been available. It is intended to provide more concise,
per-topic information that will be easier to maintain and keep up-to-date.
All documents are available on http://support.emc.com.
The following list represents the technical note best practice documents that are
available:
 EMC VPLEX Overview and General Best Practices

 EMC VPLEX SAN Connectivity
 EMC VPLEX Host Multipathing
 EMC VPLEX IP Networking
 EMC VPLEX and RecoverPoint Interoperability
 EMC VPLEX Data Mobility and Migrations
 EMC VPLEX REST API
 EMC VPLEX Upgrades

Host Multipath for VPLEX Overview

This document is designed to address host multipath configurations as they apply to
VPLEX.
The recommendations and requirements are based on general SAN best practices
and rigorous testing through E-Lab, QE and the VPLEX Performance Engineering Team.
Changes to software and hardware architecture both within VPLEX and within the
overall environment cause these recommendations and requirements to be adjusted
occasionally.
Please refer to the ESSM (EMC Simple Support Matrix) for recommendations and
requirements for multipath support statements.
Important VPLEX uses a Distributed Cache Model which is significantly different for the
purpose of multipath configurations than most arrays which typically use a Global Cache
Model. Additionally, from a host perspective, VPLEX is viewed as an active/active
architecture.
The Distributed Cache Model is designed so that each VPLEX Director has its own
independent cache. This means that if a host writes to one Director and then
performs a read against that write from another Director then the result is a cache
read miss. An array with Global Cache would result is a cache read hit. To
compensate for these differences, the host multipath configuration setting will differ
when connected to VPLEX from an array so that the chance of a cache read hit can be
optimized.
Also as mentioned, VPLEX is an active/active architecture from the host perspective.

This is completely independent from the backend array. VPLEX will handle the load
balancing and path access for the backend array and will present the device to the
host as active/active. This means that there is absolutely no consideration for
active/passive or ALUA configuration settings within the host multipath configuration
for the devices accessed through VPLEX even if the backend array where that device
resides is active/passive or ALUA.

Multipath
Software
There are several choices when selecting the host multipath software. These can be
broken down into two basic categories:
1. Path Failover
2. Load Balancing
Path Failover
Path Failover is designed as single active path and the other available paths are for
failover access only. This design has advantages with the VPLEX Distributed Cache
model as it has the highest possible chance of a cache read hit. All data will be
written and read from the same VPLEX Director for that host and if the data is in cache
on VPLEX then it will be on that one Director that the active path is configured to and
there will be a 100% chance of a cache read hit.
While this may be a distinct advantage when considering performance, there are a
few disadvantages. First, this type of configuration requires manual configuration to
assign the active path to the least utilized VPLEX Director and may create a lot of
additional work for the administrator. Second, this type of software may not have the
same responsiveness and adaptability in the event of various environment failures as
compared to load balancing software. Third, VPLEX offers a very unique type of
configuration called “Host Cross-Cluster Connect Configuration”. This is a
configuration where the host has access to the device across two different VPLEX
Clusters in a Metro Configuration across data centers. The initial configuration would
dictate that the active path be defined to the local VPLEX Cluster however if the local
VPLEX Cluster were to fail then there would be a high possibility that the first
available failover paths would be to the same Director on the remote VPLEX Cluster
for all hosts which would overwhelm that director and create a serious workload for
the administrator to try and rebalance the active paths. All this work would then have
to be redone when the original VPLEX Cluster came back online to rebalance the
active paths again on the local cluster.
Load Balancing
Load Balancing software is designed to take advantage of the additional available
paths by utilizing additional resources to move data. This increases the total
bandwidth available to move larger amounts of data faster.
Based on the specific load balancing software chosen, there are typically several
choices for configuration settings. In basic terms, these settings fall into a few
different categories:
1. Path Policy - i.e. Round Robin, Adaptive…

2. How much I/O is sent before moving on to next path
3. Path timeouts and retries
Path Policy is a setting within the multipath software and is configured at the host
level. The Path Policies vary from software vendor to software vendor and is
important to reference the ESSM (EMC Simple Support Matrix) before committing to a
product or setting.

Most products available today were designed to provide multiple path access within
the datacenter. VPLEX offers additional capabilities and therefore additional
challenges for host access.
As mentioned with the Path Failover software, VPLEX uses a Distributed Cache model
and requires multipath software to be configured to optimize the VPLEX caching
algorithms which differ from most arrays.
Not all host multipath software is created equal. Basic configurations within the
datacenter may offer the option of selecting host based native multipath solutions or
third party software but there will be circumstances where we recommend PowerPath.
PowerPath has added additional support specifically for VPLEX Metro which is not
available with any other product on the market.
Specific recommendations and settings will be discussed in the following chapters.

Chapter 2: ESXi NMP Policy
Chapter 2 ESXi NMP Policy
VMWARE ESXi ........................................................................................................ 11
Round Robin - New Policy Recommendation ........................................................... 11
Changing Policy from Fixed Path to Round Robin ................................................... 12
Additional Information........................................................................................... 12

VMWARE ESXi
For detailed information about VMWare ESXi host configuration for VPLEX, refer to the
following ELAB documentation available at http://support.EMC.com for configuration
and administration operations: https://support.emc.com/docu5265_Host-
Connectivity-Guide-for-VMWare-ESX-Server.pdf?language=en_US
Round Robin - Multipath Policy Recommendation

Round Robin This section provides the recommendations and steps to achieve optimal performance
and continuous availability with EMC® VPLEX™
 The recommended multipath setting is Round Robin for VPLEX Local and VPLEX
Metro (non-cross-connect). The I/O Limit value should be left at the default
setting of 1000.
 For VPLEX Metro cross-connect with VMware, PowerPath/VE is highly
recommended.
 PowerPath/VE 5.8 includes the auto-standby feature which allows each
ESXi host to automatically prefer to send I/O to its local VPLEX cluster
over the remote cluster. The host paths connected to the local VPLEX
Cluster will be the active paths whereas those connected to the remote
VPLEX Cluster will be the standby paths.
 For more information on PowerPath/VE and the auto-standby feature,
see the support
page:https://support.emc.com/products/1800_PowerPath-VE-for-
VMware
 Additional background: There are two problems with using NMP for VPLEX
Metro cross-connect environments:
 Round-robin path policy for a host connected to both VPLEX clusters will
incur extra read and write latency for I/O operations to the remote cluster.
Roughly half of the I/O will be local and half will be remote. WAN bandwidth
for front-end host traffic will be consumed. Additional VPLEX inter-cluster
cache-coherency traffic will be sent between clusters.
 Fixed path policy requires a lot of manual administrative work to have all
ESXi hosts and all volumes on both clusters to prefer their local cluster. For
a handful of hosts and only a few volumes this might be acceptable. But for
hundreds of hosts and thousands of volumes this is too onerous.
 In addition, should the single preferred path fail for whatever reason, the
new path chosen by a host might be at the remote cluster. And it's entirely
possible that multiple hosts could by the luck of the draw unfortunately all
choose the same new remote director and thus overload that one director.
A manual re-balancing of paths would be required at the new cluster, and
then when the old cluster is back online, the exercise has to be repeated all
over again.


Changing Policy from Fixed Path to Round Robin
Changing Policy To change the default policy setting from Fixed to Round Robin for EMC VPLEX
(Invista) devices, complete the following steps:
1. Open the vSphere CLI (recommended) or the service console.

2. Run the following command:
– For ESXi 5.x:

# esxcli storage nmp satp set --default-psp=VMW_PSP_RR --
satp=VMW_SATP_INV
– For ESXi/ESX 4.x:
# esxcli nmp satp setdefaultpsp --psp VMW_PSP_RR --satp
VMW_SATP_INV
 Set the Multipath Round Robin policy for the I/O operation limit to a value of
1000 for optimal performance with VPLEX. Currently, the VMware default value
for this setting is 1000.
1. Open the vSphere CLI (recommended) or the service console.
2. Run the following commands:
– For ESXi 5.x:

To check the I/O operations limit:
# esxcli storage nmp psp roundrobin deviceconfig get --
device=device_NAA
To set the I/O operations limit:
#esxcli storage nmp psp roundrobin deviceconfig set --
device=device_NAA --iops=1000 --type iops
– For ESXi/ESX 4.x:
To check the I/O operations limit:
# esxcli nmp roundrobin getconfig --device=device_NAA
To set the I/O operations limit:
# esxcli nmp roundrobin setconfig --device=device_NAA --iops
1000 --type iops
Additional Information
For information on VMware and VPLEX versions recommended settings, refer to
VMware Compatibility Guide for EMC VPLEX:

 http://www.vmware.com/resources/compatibility/detail.php?deviceCategory=san&pr
oductid=12738&deviceCategory=san&keyword=vplex&isSVA=1&page=1&display_i
nterval=10&sortColumn=Partner&sortOrder=Asc
For related information, refer to the following VMware Knowledgebase article:

Changing the default pathing policy for new/existing LUNs (1017760) located
at http://kb.vmware.com/kb/1017760
For information on VPLEX host cross-cluster connect configuration using PowerPath

and autostandby when using an ICL instead of ISL with the latency to each VPLEX
cluster being nearly identical please see the Knowledgebase article:
Solution: How to modify esxi powerpath paths to standby alive at boot time on esxi
host
Specific recommendations and settings will be discussed in the following chapters.
Note: Disk.AutoremoveOnPDL settings change from release ESXi 5.5 to 6.0 and may change
again in future releases. For information on Disk.AutoremoveOnPDL settings please refer to
VMWare KB article: https://kb.vmware.com/kb/2059622

Chapter 3: PowerPath
Chapter 3 PowerPath
Introduction........................................................................................................... 15
Supported Versions and Settings ........................................................................... 16

Introduction
PowerPath is a host-based software that provides path management. PowerPath
operates with several storage systems including VPLEX, on several operating
systems, with Fibre Channel.
Path management PowerPath works with VPLEX to intelligently manage the frontend I/O paths.
Path refers to the physical route between a host and VPLEX Virtual Volume or Logical
Unit (LU) as referred to by PowerPath.
This includes the host bus adapter (HBA) port, cables, a switch, VPLEX interface and
port, and an LU. LU refers to a virtual volume addressable as a single storage volume
behind a VPLEX target. For the iSCSI standard, path is the Initiator- Target-LUN, or ITL
nexus and encompasses the connection between the HBA, Storage Port and LUN.
Bus refers to two connected SAN edge points (for example, Fibre Channel fabric N-
port addresses) in the storage configuration: an HBA port on the server on one end
and an array port on the other. For the iSCSI standard, bus is the Initiator-Target, or IT
nexus. This differs from a storage path, which refers to a host's end-to-end storage
connection with an LU. Typically, multiple storage paths traverse a single bus.
PowerPath supports multiple paths to a logical device, enabling PowerPath to

provide:
 Automatic failover in the event of a hardware failure. PowerPath automatically

detects path failure and redirects I/O to another path.
 Dynamic multipath load balancing. PowerPath distributes I/O requests to a
logical device across all available paths, thus improving I/O performance and
reducing management time and downtime by eliminating the need to configure
paths statically across logical devices.
PowerPath PowerPath features include:
Features
 Multiple paths, for higher availability and performance — PowerPath supports
multiple paths between a logical device and a host. Having multiple paths
enables the host to access a logical device even if a specific path is
unavailable. Also, multiple paths can share the I/O workload to a given logical
device.
 Path management insight capabilities—PowerPath characterizes I/O patterns
and aides in diagnosing I/O problems due to flaky paths or unexpected latency
values.
Metrics are provided on:
 Read, write MB/seconds per LUN

 Latency distribution: high and low water marks per path
 Retries: number of I/Os that did not succeed down a path
PowerPath also defines and measures performance on I/O throughput, fault
detection, and path restore. Three new CLI commands (powermt set perfmon,
powermt display perf, and powermt display perf bus) provide this information. The

EMC PowerPath Family CLI and System Messages Reference Guide provides more
information. The commands associated with path management insight may cause
some performance overhead. However, EMC qualifications and in-house testing have
displayed no measurable impact to performance.
 Expanded support for standby: autostandby—An autostandby feature has been

added to the standby feature to automatically put paths into autostandby that
have intermittent I/O failures (also known as flaky paths) and/or to
automatically select autostandby for high-latency paths in VPLEX cross-
connected Metro configurations. The EMC PowerPath Family CLI and System
Messages Reference Guide provides more information.
Supported Versions and Settings

Please refer to the EMC Simple Support Matrix for latest supported versions required.
Before configuring VPLEX, complete the following on each host:
 Confirm that all necessary remediation has been completed.

This ensures that OS-specific patches and software on all hosts in the VPLEX
environment are at supported levels according to the EMC Support Matrix.
 Confirm that each host is running VPLEX-supported PowerPath version and has
at least one available path to each VPLEX fabric.
 Confirm that the load-balancing and failover policy is set to Adaptive.
PowerPath is highly recommended for the host cross-cluster connect configuration.
The load-balancing and failover policy apply to:
 ESXi
 Windows
 HP-UX
 AIX
 Linux

Chapter 4: Native MPIO
Chapter 4 Native MPIO
Introduction........................................................................................................... 18
AIX 19
HP-UX .................................................................................................................... 21
Linux ..................................................................................................................... 22
Microsoft Windows ................................................................................................ 31
Solaris ................................................................................................................... 37

Introduction
Most host operating systems provide a native form of multipath software known as
MPIO. This allows utilization of additional paths for failover or multipath access while
lowering the overhead cost involved with implementing a solution. These solutions
are relatively basic and designed for solutions within the data center.
VPLEX has tested and qualified these solutions and provide guidance found in the
ELAB Host Configuration Guides.
The following sections summarize the MPIO configuration settings specific to VPLEX.

IBM AIX
For detailed information about IBM AIX host configuration for VPLEX, refer to the
Connectivity-Guide-for-IBM-AIX.pdf?language=en_US
Before configuring VPLEX in the IBM AIX environment, complete the following on each
host:
 Confirm that all necessary remediation has been completed. This ensures that
OS-specific patches and software on all hosts in the VPLEX environment are at
supported levels according to the EMC Support Matrix.
 Confirm that each host is running VPLEX-supported failover software and has at
least one available path to each VPLEX fabric.
Note: Always refer to the EMC Support Matrix for the most up-to-date support information
and prerequisites.
 If a host is running EMC PowerPath, confirm that the load-balancing and

failover policy is set to Adaptive.
 Support requires minimum AIX 7100-00-02-1041, AIX 6100-04-03-1009, and
AIX 5300-11-02-1009.
 AIX attach to VPLEX requires minimum EMC ODM 5.3.0.3.
 PowerPath, Veritas DMP, and AIX MPIO are the only multipath solutions
supported with VPLEX and AIX hosts.
 AIX MPIO minimum requirements are as follows:
 Use ODM files on each host. This includes the reset_delay parameter set to
0. The ODM versions are posted at
ftp://ftp.emc.com/pub/elab/aix/ODM_DEFINITIONS
 AIX 6.1 Technology Level 6100-08-00-1241, 6100-07-06-1241 and
later. Requires minimum ODM version 5.3.0.8
 AIX 7.1 Technology Level 7100-02-00-1241, 7100-01-06-1241 and
later. Requires minimum ODM version 6.0.0.3
 VIOS 2.2.2.0 and later. Requires minimum ODM version 5.3.0.8
 VIOS 2.2.2.1-SP1 FP26 and later. Requires minimum ODM version
5.3.0.8
The host must be rebooted after updating ODM file for the settings to take effect.
 The following configurations with AIX 6.1 and 7.1 are supported with MPIO:
 AIX LVM must be used
 IBM PowerHA, GPFS filesystem
 VIOS, VIOC, LPAR, NPIV

 The following configurations are not supported at this time:

 Oracle RAC
 Veritas VxVM, VCS, SFRAC, VxCFS
 VPLEX 5.1.x is the minimum version supported
IMPORTANT
For optimal performance in an application or database environment, ensure alignment of

your host's operating system partitions to a 32 KB block boundary.
Veritas DMP settings with VPLEX

 Veritas DMP 5.1 SP1 requires the asl package 5.1.100.100 to correctly detect
VPLEX
 If a host attached to VPLEX is running Veritas DMP multipathing, change the
following values of the DMP tunable parameters on the host to improve the way
DMP handles transient errors at VPLEX in certain failure scenarios:
 dmp_lun_retry_timeout for the VPLEX array to 60 seconds using the
following command:
"vxdmpadm setattr enclosure emc-vplex0 dmp_lun_retry_timeout=60"
 recoveryoption to throttle and iotimeout to 30 using the following

command:
"vxdmpadm setattr enclosure emc-vplex0 recoveryoption=throttle iotimeout=30"

HP-UX
Note: please reference -
Before configuring VPLEX in the HP-UX environment, complete the following on each
host:
 Confirm that all necessary remediation has been completed

and prerequisites.

failover policy is set to Adaptive
 The following are supported with VPLEX:
 HPUX 11iv3 with VxVM 5.1 SP1
 HP-UX 11iv3 with VxMP 5.0 - 6.0
 HP-UX 11iv2 with VxMP 4.1 - 5.0
IMPORTANT

Veritas DMP settings with VPLEX

 Veritas DMP 5.1 SP1 requires the asl package 5.1.100.100 to correctly detect
the VPLEX Virtual Volumes
 If a host attached to VPLEX is running Veritas DMP multipathing, change the
following values of the DMP tunable parameters on the host to improve the way
DMP handles transient errors at the VPLEX array in certain failure scenarios:
following command:
"vxdmpadm setattr enclosure emc-vplex0 dmp_lun_retry_timeout=60"
 recoveryoption to throttle and iotimeout to 30 using the following

command:
"vxdmpadm setattr enclosure emc-vplex0 recoveryoption=throttle iotimeout=30"

Linux
For detailed information about Linux host configuration for VPLEX, refer to the
Connectivity-Guide-for-Linux.pdf?language=en_US
Native MPIO
DM-MPIO, native multipathing, provides a mechanism to address device names
persistently through the use of udev and scsi-id. The names used to address
multipath names rely on the properties of the physical device, and are thus both
unique and consistent across reboots.
Red Hat Enterprise The following steps detail the procedure for configuring native multipath failover for
Linux (RHEL) Invista or VPLEX virtualized storage in a RHEL 4 host.
RHEL 4
To configure MPIO on a RHEL 4 host, complete the following steps:
1. Install the multipath-tools rpm package.

The device-mapper-multipath-tools package is not installed as part of a default
installation of the operating system. Select the package as part of the OS install or
install the rpm later either from the install CD or by downloading the rpm from the Red
Hat website.
2. Verify that the required version of udev has been installed for your
environment. To query the udev version, execute the following command:
# rpm -q udev
If required, upgrade the udev package and then execute the command to create a
new device under the ’/dev’ directory.
# udevstart
3. Load the dm_multipath kernel module if it is not already loaded, as follows:

# modprobe dm_multipath
4. Replace the default /etc/multipath.conf with the following multipath.conf file

recommended by EMC for attach to EMC storage.
To find the file on a RHEL system, issue the following command:
# rpm -ql `rpm -qa | grep multipath` | grep synt
The following is sample output from RHEL:
/usr/share/doc/device-mapper-multipath-0.4.5/multipath.conf.synthetic

IMPORTANT
The following is only an example. Consult Red Hat's documentation to ensure the correct
syntax is followed for your release. MPIO continues to evolve with each release of RHEL.
Note: VPLEX require its own specific configuration stanza in /etc/miltipath.conf file.
## This is the /etc/multipath.conf file recommended for

## EMC storage devices.
##
## OS: RHEL 4
## Arrays: Invista/VPLEX
##
## The blacklist is the enumeration of all devices that are to be
## excluded from multipath control
blacklist {
## Replace the wwid with the output of the command MPIO
## 'scsi_id -g -u -s /block/[internal scsi disk name]'
## Enumerate the wwid for all internal scsi disks.
##
wwid 35005076718 d4224d
devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"
devnode "^hd[a-z][0-9]*"
devnode "^cciss!c[0-9]d[0-9]*[p[0-9]*]"
}
## Use user friendly names, instead of using WWIDs as names.
defaults {
user_friendly_names yes
}
devices {
# Device attributes requirements for Invista/VPLEX
device {
vendor "EMC"
product "Invista"
path_checker tur
no_path_retry 5
product_blacklist "LUNZ"
}
}
5. Perform a dry run and evaluate the setup by running the multipath command.
# multipath -v2 -d
create: mpath42 (360060480000190100501533030383644)
[size=5 GB][features="0"][hwhandler="0"]
\_ round-robin 0
\_ 11:0:1:52 sdao 66:128
\_ 11:0:2:52 sdaz 67:48
\_ 10:0:1:52 sdm 8:192
\_ 10:0:2:52 sdx 65:112
create: mpath43 (360060480000190100501533030383645)
[size=5 GB][features="0"][hwhandler="0"]
\_ round-robin 0
\_ 11:0:1:53 sdap 66:144
\_ 11:0:2:53 sdba 67:64
\_ 10:0:1:53 sdn 8:208
\_ 10:0:2:53 sdy 65:128
6. If the listing is appropriate, commit the configuration as follows:

a. Start the required multipath processes.
# /etc/init.d/multipathd start

b. Execute the multipath command.

# multipath -v2
c. Perform an lsmod and verify that the processes are running.
# lsmod |grep dm
dm_round_robin 4929 1
dm_multipath 22097 2 dm_round_robin
dm_mod 66433 1 dm_multipath
7. To get a listing of the current setup do:

# multipath -ll
8. Integrate the startup of the appropriate daemons in the boot sequence as

follows:
# chkconfig --add multipath
# chkconfig multipathd on
You may optionally reboot the host to verify that the required processes automatically
startup.
Additional documentation is available in the directory /usr/share/doc/device-

mapper-multipath-x.x.x/, the manpages on your system, and from Red Hat's website.
RHEL 5 and RHEL 6

The following steps detail the procedure for configuring native multipath failover on a
RHEL 5 or RHEL 6 host.
The /etc/multipath.conf file is installed by default when you install the device-
mapper-multipath package.
To configure MPIO on a RHEL 5 or RHEL 6 host, complete the following steps:
1. Replace the default /etc/multipath.conf with the following multipath.conf file

recommended by EMC for attach to EMC storage.
To find the file on a RHEL system to modify, issue the following command:
The following is a sample output from RHEL:
/usr/share/doc/device-mapper-multipath-0.4.5/multipath.conf.synthetic
IMPORTANT
The following is only an example. Consult Red Hat documentation to ensure the correct
syntax is followed for your release. MPIO continues to evolve with each release of RHEL.
The Linux native MPIO in RHEL 5.5 and later and RHEL 6.0 and later already contain
default configuration parameters for EMC VPLEX virtualized storage to provide

optimal performance in most environments. There is no need to create a device

stanza for VPLEX unless you wish to modify the default behavior.
If you wish to modify the default behavior the default stanza is contained in the
multipath.conf.defaults located in usr/share/doc/device-mapper-multipath-x.x.x. The
device stanza begins with the following:
vendor "EMC"
product"Invista"
Copy this stanza into your /etc/multipath.conf file and modify it as desired.
# This is an example configuration file for device mapper multipath.
# For a complete list of the default configuration values, see
# /usr/share/doc/device-mapper-multipath-0.4.5/multipath.conf.defaults
# For a list of configuration options with descriptions, see
# /usr/share/doc/device-mapper-multipath-0.4.5/multipath.conf.annotated
# Blacklist all devices by default. Remove this to enable multipathing
# on the default devices.
Note: Insert # to disable the all device blacklist and enable multipathing on your system.
#blacklist {
# devnode "*"
#}
## By default, devices with vendor = "IBM" and product = "S/390.*" are
## blacklisted. To enable multipathing on these devices, uncomment the
## following lines.
#blacklist_exceptions {
# device {
# vendor "IBM"
# product "S/390.*"
# }
#}
defaults {
}
##
## This is a template multipath-tools configuration file
## Uncomment the lines relevant to your environment
230
Native Multipath Failover
##
#defaults {
# udev_dir /dev
# polling_interval 10
# selector "round-robin 0"
# path_grouping_policy multibus
# getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
# prio_callout /bin/true
# path_checker readsector0
# rr_min_io 100
# rr_weight priorities
# failback immediate
# no_path_retry fail
# user_friendly_name yes
#}
##
## The wwid line in the following blacklist section is shown as an
example
## of how to blacklist devices by wwid. The 3 devnode lines are the
## compiled in default blacklist. If you want to blacklist entire types
## of devices, such as all scsi devices, you should use a devnode line.
## However, if you want to blacklist specific devices, you should use
## a wwid line. Since there is no guarantee that a specific device will
## not change names on reboot (from /dev/sda to /dev/sdb for example)
## devnode lines are not recommended for blacklisting specific devices.

##
blacklist {
wwid 360060480000190101965533030303230
devnode "^hd[a-z]"
devnode "^cciss!c[0-9]d[0-9]*"
}
#multipaths {
# multipath {
# wwid 3600508b4000156d700012000000b0000
# alias yellow
# path_selector "round-robin 0"
# failback manual
# no_path_retry 5
# }
# multipath {
# wwid 1DEC_____321816758474
# alias red
# }
#}
#devices {
# device {
# vendor "COMPAQ "
# product "HSV110 (C)COMPAQ"
# getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
# path_selector "round-robin 0"
# hardware_handler "0"
# failback 15
# no_path_retry queue
# }
# device {
# vendor "COMPAQ "
# product "MSA1000 "
# }
#}
devices {
# Device attributes for EMC Invista/VPLEX
device {
vendor "EMC"
product "Invista"
path_checker tur
no_path_retry 5
}
}
2. Perform a dry run and evaluate the setup by running the multipath command:
# multipath -v2 -d [RHEL 5 & 6]
create: mpath15 (360060480000190101965533030423744) EMC,INVISTA
[size=8.4G][features=0][hwhandler=0]
\_ round-robin 0 [prio=2][undef]
\_ 2:0:0:49 sdp 8:240 [undef][ready]
\_ 3:0:0:49 sds 65:32 [undef][ready]
[size=8.4G][features=0][hwhandler=0]
\_ 2:0:0:50 sdq 65:0 [undef][ready]
\_ 3:0:0:50 sdt 65:48 [undef][ready]
3. If the listing is appropriate, commit the configuration as follows:


b. Execute the multipath command:

# multipath -v2
c. Perform an lsmod and verify that the processes are running.
With VPLEX attached, the following modules will be listed:
# lsmod |grep dm
4. To get a listing of the current setup issue the following command:
# multipath -ll
follows:
# chkconfig --add multipathd
# chkconfig multipathd on
startup.
Note: Additional documentation is available in the directory /usr/share/doc/device-mapper-

multipath-x.x.x/.
SuSE Linux This section details the procedure for configuring native multipath failover (MPIO) for
Enterprise Server EMC VPLEX virtualized storage on a SuSE Linux Enterprise Server (SLES) host.
(SLES)
SLES 9 and SLES 10
The following steps apply for SLES 9 and SLES 10 servers.
To configure MPIO on a SLES host, complete the following steps:
1. Install the multipath-tools rpm package.

The multipath-tools package is not installed as part of a default installation of the
operating system. The recommended multipath-tools rpm package is available in the
installation medium for the operating system or from the SuSE website.
2. Verify that the required version of udev has been installed for your
environment. To query the udev version, execute the following command:
# rpm -q udev
If required, upgrade the udev package and then recreate the devices.
 On SLES 9 execute:
# udevstart
 On SLES 10 execute:
# /etc/init.d/boot.udev restart
3. On SLES 9, SuSE recommends that 'subfs' auto-mount be disabled when

using MPIO. This is achieved by editing the '/etc/sysconfig/hotplug' and
setting the value, as follows:
HOTPLUG_USE_SUBFS=no

Also, to prevent significant boot delays, edit /etc/sysconfig/boot and set the value as
follows:
DISABLE_BLKID=yes
4. Load the dm_multipath kernel module if it is not already loaded, as follows:

# modprobe dm_multipath
5. The following is the /etc/multipath.conf file recommended for attach to EMC

storage.
If /etc/multipath should not exist on your server, you can locate a copy to use by
using the following Linux command:
The following is sample output from SLES 9:

/usr/share/doc/packages/multipath-tools/multipath.conf.synthetic
IMPORTANT
The following is only an example. Consult SuSE documentation to ensure the correct syntax
is followed for your release. MPIO continues to evolve with each release of SLES.
Note: EMC VPLEX require its own specific configuration stanza in /etc/multipath.conf file.
 On a SLES 9 system:
## EMC storage devices.
##
## OS : SLES 9
## Arrays : INVISTA/VPLEX
##
## The blacklist is the enumeration of all devices that are to be
## excluded from multipath control
blacklist {
## Replace the wwid with the output of the command
##
wwid 35005076718d4224d
devnode "^hd[a-z][[0-9]*]"
}
defaults {
}
devices {
## Device attributes for EMC SYMMETRIX
device {
vendor "EMC "
product "INVISTA"
path_grouping_policy multibus
path_checker tur
no_path_retry 5
}
}

 On a SLES 10 SP3 and later and SLES 11 and later systems, the Linux native
MPIO already contains default configuration parameters for EMC VPLEX
virtualized storage to provide optimal performance in most environments. There
is no need to create a device stanza for VPLEX unless you wish to modify the
default behavior.
IMPORTANT
The following is an example of an edited file that would support VPLEX. Consult SuSE
documentation to ensure the correct syntax is followed for your release. MPIO continues to
evolve with each release of SLES.
If you wish to modify the default behavior the default stanza is contained in the
multipath.conf.defaults located in usr/share/doc/device-mapper-multipath-x.x.x. The
device stanza begins with the following:
vendor "EMC"
product"INVISTA"
Copy this stanza into your /etc/multipath.conf file and modify it as desired.
## EMC storage devices
##
## OS : SLES 10
## Arrays: INVISTA/VPLEX
##
##
defaults {
}
blacklist {
## Replace the wwid with the output of the command MPIO
##
wwid 35005076718d4224d
devnode "^hd[a-z][0-9]*"
}
devices {
## Device attributes for EMC SYMMETRIX
device {
vendor "EMC "
product "INVISTA"
path_grouping_policy multibus
path_checker tur
no_path_retry 5
}
}
6. Perform a dry run and evaluate the setup by running the multipath command:
 With Invista or VPLEX virtualized storage logical units on a SLES 9 host, the
output will look similar to the following:
# multipath –v2 -d
create: 360060480000190100501533030383842
[size=898 MB][features="0"][hwhandler="0"]
\_ round-robin 0 [prio=4]
\_ 2:0:2:105 sdaa 65:160 [ready]

\_ 2:0:3:105 sdas 66:192 [ready]

\_ 3:0:2:105 sdcj 69:112 [ready]
\_ 3:0:3:105 sddb 70:144 [ready]
create: 360060480000190100501533030383843
[size=898 MB][features="0"][hwhandler="0"]
\_ round-robin 0 [prio=4]
\_ 2:0:2:106 sdab 65:176 [ready]
\_ 2:0:3:106 sdat 66:208 [ready]
\_ 3:0:2:106 sdck 69:128 [ready]
\_ 3:0:3:106 sddc 70:160 [ready]
 With Invista or VPLEX virtualized storage logical units on a SLES 10 host, the
output will look similar to the following:
# multipath –v2 –d
[size=468M][features=0][hwhandler=0]
\_ 11:0:0:39 sdbr 68:80 [undef][ready]
\_ 11:0:1:39 sdcc 69:0 [undef][ready]
\_ 10:0:0:39 sdl 8:176 [undef][ready]
\_ 10:0:1:39 sdw 65:96 [undef][ready]
[size=468M][features=0][hwhandler=0]
\_ 11:0:0:40 sdbs 68:96 [undef][ready]
\_ 11:0:1:40 sdcd 69:16 [undef][ready]
\_ 10:0:0:40 sdm 8:192 [undef][ready]
\_ 10:0:1:40 sdx 65:112 [undef][ready]
7. If the listing is appropriate, commit the configuration as follows.

# /etc/init.d/boot.multipath start
b. Perform an lsmod and verify that the processes are running.

# lsmod |grep dm
8. To get a listing of the current setup, type:

# multipath -ll

follows:
# insserv boot.device-mapper multipathd boot.multipath
startup.
Note: Additional documentation is available in the directory

/usr/share/doc/packages/multipath-tools/x.x.x/.

Microsoft Windows
For detailed information about Microsoft Windows host configuration for VPLEX, refer
to the following ELAB documentation available at http://support.EMC.com for
configuration and administration operations:
https://support.emc.com/docu5134_Host-Connectivity-Guide-for-
Windows.pdf?language=en_US
Important
VPLEX has dropped support for Windows 2003.
Always refer to the VPLEX ESSM for support details located on

http://support.emc.com
Note: Native MPIO is not supported with failover clusters when stretched, or cross-
connected. PowerPath must be used. See EMC knowledgebase article emc000187975.
Windows 2008
Server Core and
Windows 2008 R2 Note: Refer to Microsoft documentation for installing the Microsoft Multipath I/O feature.
Server Core
Native MPIO must be configured to manage VPLEX. This can be accomplished by
opening Control Panel then selecting the MPIO applet.
The claiming of array/device families can be done in one of two ways as described in
“Method 1”, and in “Method 2”.
Method 1
Manually enter the Vendor and Device IDs of VPLEX for native MPIO to claim and
manage.
Note: This may be the preferred method if all arrays are not initially connected during
configuration and subsequent reboots are to be avoided.
To manually enter the array vendor and product ID information:
1. Use the MPIO-ed Devices tab in the MPIO Properties control panel applet.
2. Select Add and enter the vendor and product IDs of the VPLEX devices to be
claimed by native MPIO.
The vendor ID must be entered as a string of eight characters (padded with trailing
spaces) and followed by the product ID entered as a string of sixteen characters
(padded with trailing spaces).
For example, to claim a VNX series and CLARiiON RAID 1 device in MPIO, the string
would be entered as:
EMC*****Invista*********
The asterisk is representative of a space.

Figure 1 Array and device types
Method 2
Use the MPIO applet to discover, claim, and manage the arrays already connected
during configuration.
Note: This may be the preferred method if ease-of-use is required and subsequent reboots
are acceptable when each array is connected.
IMPORTANT
MPIO limits the number of paths per LUN to 32. Exceeding this number will result in the host
crashing with a Blue Screen stop message. Do not exceed 32 paths per LUN when
configuring MPIO on your system.
Automatic discovery is configured using the Discover Multi-Paths tab of the MPIO
Properties control panel applet. Note that only arrays which are connected with at
least two logical paths will be listed as available to be added in this tab, as follows:
 Devices from VPLEX arrays will be listed in the Others section of the applet
Select the array / device types to be claimed and managed by MPIO by selecting the
Device Hardware ID, and clicking the Add button.
Note: The OS will prompt you to reboot for each device type added. A single reboot will
suffice after multiple devices types are added.
Path management Following reboot, after all device types have been claimed by MPIO, each VPLEX-
in Multipath I/O based disk will be shown in Device Manager as a Multi-Path Disk Device. When
for VPLEX managed by MPIO, a new tab, named MPIO, will be available under Properties of the
selected disk device. Under the MPIO tab, the number of logical paths configured
between the host and array should be reported.
The default Load Balance Policy (as reported in the MPIO tab) for each disk device
depends upon the type of disk device presented.
VPLEX devices will report a default Load Balance Policy as "Round Robin" with all
active paths as "Active/Optimized". The default policy can be overridden by changing
the Load Balance Policy to any available, except "Fail Over Only". See the Windows
Server 2008 and Windows Server 2008 R2 documentation for a detailed description
of available Load Balance policies.
Note: The default Load Balance Policy cannot be changed globally for all disk devices. The
change must be done on a per-disk device basis.
Enabling Native Windows Server 2008 Server Core and Windows Server 2008 R2 Server Core
MPIO MPIO and other features must be started from the command line since Windows
Server 2008 Server Core and Windows Server 2008 R2 Server Core are minimal

installations that do not have traditional GUI interfaces. Refer to

http://technet.microsoft.com for more information on Windows Server Core
installations.
To enable the native MPIO feature from the command line, type:
start /w ocsetup MultipathIo
After the system reboots, you can manage MPIO with the mpiocpl.exe utility. From the
command prompt, type:
mpiocpl.exe
The MPIO Properties window displays. From here, arrays/devices can be claimed and
managed as described in the section above for standard Windows installations.
For more information on Microsoft Native MPIO, refer to:
http://www.microsoft.com and http://technet.microsoft.com.
Native MPIO with Support for Native MPIO in Windows Server 2012
Windows Server Windows Server 2012 include native multipathing (MPIO) support as a feature of the
2012 OS.
Native MPIO is supported with EMC CX4 Series, DMX-4, and VMAX storage array
models.
Note the following:
 To use the Microsoft default DSM, storage must be compliant with SCSI Primary
Commands-3 (SPC-3)
 Default Microsoft MPIO Timer Counters are supported
 VPLEX is active/active and does not support mode 4 [ALUA]
Configuring Native This section explains how to configure native MPIO for EMC storage arrays. Native
MPIO for Windows MPIO is installed as an optional feature of the Windows Server 2012.
Server 2012
Note: Refer to Microsoft documentation for installing the Microsoft Multipath I/O feature.
Configuring MPIO and installing DSM

When MPIO is installed, the Microsoft device-specific module (DSM) is also installed,
as well as an MPIO control panel. The control panel can be used to do the following:
 Configure MPIO functionality

 Install additional storage DSMs
 Create MPIO configuration reports
Opening the MPIO control panel
Open the MPIO control panel either by using the Windows Server 2012 control panel
or by using Administrative Tools.

To open the MPIO control panel using the Windows Server 2012 control panel,
complete the following steps:
1. On the Windows Server 2012 desktop, move your mouse to the lower left
corner and click Start.
2. Click MPIO.
To open the MPIO control panel using Administrative Tools, complete the following
steps:
1. On the Windows Server 2012 desktop, move your mouse to the lower left
corner and click Start.
2. Point to Administrative Tools and click MPIO.
The MPIO control panel opens to the Properties dialog box.
Note: To access the MPIO control panel on Server Core installations, open a command
prompt and type MPIOCPL.EXE.
Once installed, native MPIO must be configured to manage VPLEX. Open Control
Panel, then the MPIO applet.
Device discovery IMPORTANT

and claiming
devices for MPIO MPIO limits the number of paths per LUN to 32. Exceeding this number will result in the host
crashing with a Blue Screen stop message. Do not exceed 32 paths per LUN when
configuring MPIO on your system.
Automatic discovery is configured using the Discover Multi-Paths tab of the MPIO
Properties control panel applet. Note that only arrays which are connected with at
least two logical paths will be listed as available to be added in this tab, as follows:
 Devices from VPLEX arrays will be listed in the Others section of the applet
Select the array and device types to be claimed and managed by MPIO by selecting
the Device Hardware ID, and clicking Add.

Figure 2 MPIO Properties dialog box
Note: The OS will prompt you to reboot for each device type added. A single reboot will
suffice after multiple devices types are added.
Path management Following reboot, after all device types have been claimed by MPIO, each VPLEX-
in Multipath I/O based disk will be shown in Device Manager as a Multi-Path Disk Device.
for VPLEX
When managed by MPIO, a new tab, MPIO, will be available under Properties of the
selected disk device. Under the MPIO tab, the number of logical paths configured
between the host and array should be reported.
This tab will also allow you to change the MPIO load balancing policy for a disk
device.
Note: Some load balancing policies may not be available for specific array disk types.
For example, the Round Robin policy is not available for VNX disk devices, but Round
Robin with Subset is.
Options available for load balancing policies are as follows:
 Fail Over Only — Policy that does not perform load balancing. This policy uses a
single active path, and the rest of the paths are standby paths. The active path
is used for sending all I/O. If the active path fails, then one of the standby
paths is used. When the path that failed is reactivated or reconnected, the
standby path that was activated returns to standby.
 Round Robin — Load balancing policy that allows the Device Specific Module
(DSM) to use all available paths for MPIO in a balanced way. This is the default
policy that is chosen when the storage controller follows the active-active

model and the management application does not specifically choose a load-
balancing policy.
 Round Robin with Subset — Load balancing policy that allows the application
to specify a set of paths to be used in a round robin fashion, and with a set of
standby paths. The DSM uses paths from a primary pool of paths for processing
requests as long as at least one of the paths is available. The DSM uses a
standby path only when all the primary paths fail. For example, given 4 paths:
A, B, C, and D, paths A, B, and C are listed as primary paths and D is the
standby path. The DSM chooses a path from A, B, and C in round robin fashion
as long as at leastone of them is available. If all three paths fail, the DSM uses
D, the standby path. If paths A, B, or C become available, the DSM stops using
path D and switches to the available paths among A, B, and C.
 Least Queue Depth — Load balancing policy that sends I/O down the path with
the fewest currently outstanding I/O requests. For example, consider that there
is one I/O that is sent to LUN 1 on Path 1, and the other I/O is sent to LUN 2 on
Path 1. The cumulative outstanding I/O on Path 1 is 2, and on Path 2, it is 0.
Therefore, the next I/O for either LUN will process on Path 2.
 Weighted Paths - Load balancing policy that assigns a weight to each path. The
weight indicates the relative priority of a given path. The larger the number, the
lower ranked the priority. The DSM chooses the least-weighted path from
among the available paths.
 Least Blocks - Load balancing policy that sends I/O down the path with the
least number of data blocks currently being processed. For example, consider
that there are two I/Os: one is 10 bytes and the other is 20 bytes. Both are in
process on Path 1, and both have completed Path 2. The cumulative
outstanding amount of I/O on Path 1 is 30 bytes. On Path 2, it is 0. Therefore,
the next I/O will process on Path 2.
The default Load Balance Policy (as reported in the MPIO tab) for each disk device
depends upon the type of disk device presented:
 VPLEX devices will report a default Load Balance Policy as "Round Robin" with
all active paths as "Active/Optimized."
Load balancing policies should be changed based on your particular environment. In
most cases, the default policy will be suitable for your I/O load needs. However,
some environments may require a change to the load balancing policy to improve
performance or better spread I/O load across storage front-end ports. EMC does not
require a specific load balancing policy for any environment, and our customers are
free to make changes to their load balancing policies as they see fit to meet their
environment's needs.
For more information on Microsoft Native MPIO, refer to:
http://www.microsoft.com and http://technet.microsoft.com

Solaris
Before configuring VPLEX in the Solaris environment, complete the following on each
host:
 Confirm that all necessary remediation has been completed.

and prerequisites.

failover policy is set to Adaptive.
 To run DMP, VPLEX 4.2 and Symantec 5.1 with the appropriate asl package are
required.
To improve the way DMP handles transient errors at the VPLEX array in certain failure
scenarios, the following attributes must be changed from the default value:

following command:
vxdmpadm setattr enclosure emc-vplex0 dmp_lun_retry_timeout=60
 recoveryoption to throttle and iotimeout to 30 using the

following command:
vxdmpadm setattr enclosure emc-vplex0 recoveryoption=throttle iotimeout=30
IMPORTANT

VPLEX and VPLEX can work with PowerPath, DMP, or MPxIO as multipath software in Oracle
MultiPath software Solaris environment, as described briefly in this section.
VPLEX and EMC PowerPath

EMC PowerPath can work with VPLEX devices as a multipath software in Oracle
Solaris host as it does with Symmetrix or CLARiiON devices in Active/Active mode.
Refer to VPLEX EMC Simple Support Matrix for supported PowerPath versions
Refer to https://support.emc.com for the PowerPath's user guide and other

PowerPath support documentation.

VPLEX and DMP

Start with VPLEX 4.2 and Symantec VxVM 5.1. You can use DMP as multipath software
in Oracle Solaris host to manage VPLEX devices.
IMPORTANT
Symantec DMP will be disabled automatically when PowerPath is installed. Uninstall

PowerPath is required if you want to use DMP as host multipath management application.
To use DMP with VPLEX devices, It required that we must install an addition ASL
package from Symantec.
Symantec ASL package for VPLEX can be downloaded from following location:
https://sort.symantec.com/asl
Follow the instruction to install ASL as provided by Symantec.
The default path policy is balanced.
The following are commands to verify ASL installation and the version of ASL.
1. Enable vxdctl and then run the vxddladm list support command to ensure that
the ASL is correctly installed.
bash-2.05# vxddladm listsupport libname=libvxInvista.so
bash-2.05# vxddladm listsupport libname=libvxInvista.so

ATTR_NAME ATTR_VALUE
==================================================================
LIBNAME libvxInvista.so
VID EMC
PID Invista
ARRAY_TYPE Inv-A/A, VPLEX-A/A
ARRAY_NAME EMC_Invista, EMC_VPLEX
2. Verify the ASL version.

bash-2.05# vxddladm listversion libname=libvxInvista.so
LIB_NAME ASL_VERSION Min. VXVM version
===================================================================
libvxInvista.so vm-5.1-rev-1 5.1
3. After installing ASL package, a VPLEX device is recognized as follows:

# vxdmpadm list dmpnode
dmpdev = emc-vplex0_00b6
state = enabled
enclosure = emc-vplex0
cab-sno = FNM00094900286
asl = libvxInvista.so
vid = EMC
pid = Invista
array-name = EMC-VPLEX
array-type = VPLEX-A/A
iopolicy = Balanced
avid = 00B6
lun-sno = 6000144000000010A001CAAF895253B6
udid =
EMC%5FInvista%5FFNM00094900286%20%20%5F6000144000000010A001CAAF895253B6
dev-attr = -
###path = name state type transport ctlr hwpath aportID aportWWN attr

path = c2t500014425001CA10d2s2 enabled(a) - FC c2

/pci@1e,600000/fibre-channel@2/fp@0,0 - 50:00:14:42:50:01:ca:10 -
path = c4t500014424001CA10d2s2 disabled - FC c4
/pci@1e,600000/fibre-channel@3/fp@0,0 - 50:00:14:42:40:01:ca:10 -
IMPORTANT
ASL is tightly coupled with the VPLEX SCSI personality. If there is any change in the SCSI
personality in the future VPLEX releases, then there should be a corresponding changes in
the ASL as applicable.
VPLEX and MPxIO Beginning with S10u11 and S11.1, you can use the MPxIO as multipath software in
Oracle Solaris host to manage VPLEX devices.
Required patches for MPxIO

To use MPxIO with VPLEX devices, the following patches are required:
 For Solaris 11.1 (both Sparc and x86):

 SRU 5.5
 For Solaris S10u11 (1/13) Sparc platform:
 148888-02 (kernel patch)
 142088-03 (fp/luxadm patch)
 150115-02 (sd patch)
 For Solaris S10u11 (1/13) x86 platform:
 148889-02 (kernel patch)
 142089-03 (fp/luxadm patch)
 150116-02 (sd patch)
Refer the README file that came with your SRU or patches for detail how to install
those patches.
Enabling MPxIO
After the patches are installed, complete the following changes in the host to enable
MPxIO:
 With S11.1 OS
 Set the parameter mpxio_disable in the file
/etc/driver/drv/fp.conf to mpxio-disable="no"
 With S10u11 OS
 Set the parameter mpxio_disable in the file
/kernel/drv/fp.conf to mpxio-disable="no"
 You must set the following parameter in the file

/kernel/drv/scsi_vhci.conf as:
load-balance="round-robin"

auto-failback="enable"
 If you are booting from the internal disk:, reboot the host.
 If you are booting from the SAN, issue the following command from the host:
'/usr/sbin/stmsboot -e'
Stmsboot enables MPxIO and makes the appropriate changes in your system file
(/etc/vfstab) so that you will be able to reboot after MPxIO is enabled since your boot
disk's name will change to MPxIO name after the reboot.
If you do not use the stmsboot command, the system may fail to reboot after MPxIO is
enabled.
VPLEX with Oracle There is a limited support of Solaris Cluster with VPLEX when using MPxIO as
Solaris Cluster multipathing.
support
IMPORTANT
MPxIO is the only multipath software that can be used with Solaris Cluster and VPLEX.
PowerPath is not supported.
To use Solaris Cluster with VPLEX and MPxIO the following restrictions must be strictly
applied:
 Support only includes Solaris 10 1/13 (S10u11) with the Solaris Cluster SC3.3
3/13 (u2) or SC3.3, SC3.3 5/11 (u1) with the latest cluster core patch and
agents patches.
 Only use cluster with local VPLEX configuration. (Stretch cluster over Metro
configuration, Campus Cluster is not supported at this time.)
 Only use cluster with no more than two (2) nodes.
 Each VPLEX devices used by the cluster must not have more than two access
paths cluster-wide.
 The quorum device must be a VPLEX device. No other shared storage device nor
Quorum Server can be used.
 The cluster must use the SCSI-2 reservation.
The cluster must be configured with the following setting (to make sure they are using
SCSI-2 reservation):
a. Cluster’s global_fencing property is set to ‘pathcount’ .

b. The quorum ‘Access Mode’ must be ‘scsi2’.
c. The property ‘default_fencing’ must set to ‘global’ for the quorum device
and all data devices.
 To check the current setting of the global_fencing property, use the
following command:
‘ /usr/cluster/bin/cluster show –t global ’

 To change the current setting of the global_fencing propery to ‘pathcount’,

use the following command:
‘ /usr/cluster/bin/cluster set –p global_fencing=pathcount ‘
 To check the current setting of ‘default_fencing’ property and the number of

device access paths for each device, use the following command:
or:
N N>
To change the ‘default_fencing’ property to ‘global’, use the following command:
---p default_fencing=global d<N
(N is the device’s did number)
 After setting your global_fencing property to ‘pathcount’. The following steps

must be made to guarantee your quorum device is using SCSI-2 reservation.
 Make sure your quorum devices is using ‘Access Mode’ is scsi2, using the
command
‘ /usr/cluster/bin/clquorum show ‘
 If the quorum device “Access Mode” is scsi3, the following steps must be
done (at least step a and step b):
a. Replace the quorum device with a new one.
b. Change the ‘default_fencing’ property of the old quorum device to
‘global’.
c. Replace the original quorum device back (in case you want to re-use the
original quorum device).


EMC® VPLEX™ Host Multipathing

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

EMC® VPLEX™ Host Multipathing

Caricato da

Copyright:

Formati disponibili

Implementation Planning and Best Practices Technical Notes

EMC® VPLEX™ Host Multipathing

Published September 2016

The information in this publication is provided as is. EMC Corporation makes no

EMC VPLEX Host Multipathing

Part Number H13547

2 EMC VPLEX Host Multipathing

Chapter 2 ESXi NMP Policy 10

Chapter 4 Native MPIO 17

EMC VPLEX Host Multipathing 3

Path management in Multipath I/O for VPLEX ...................................................... 35

4 EMC VPLEX Host Multipathing

This chapter presents the following topics:

Document Organization ........................................................................................... 6

Host Multipath for VPLEX Overview .......................................................................... 7

EMC VPLEX Host Multipathing 5

 SAN technology and network design

All documents are available on http://support.emc.com.

 EMC VPLEX Overview and General Best Practices

6 EMC VPLEX Host Multipathing

Host Multipath for VPLEX Overview

Also as mentioned, VPLEX is an active/active architecture from the host perspective.

EMC VPLEX Host Multipathing 7

1. Path Policy - i.e. Round Robin, Adaptive…

8 EMC VPLEX Host Multipathing

Specific recommendations and settings will be discussed in the following chapters.

EMC VPLEX Host Multipathing 9

Chapter 2 ESXi NMP Policy

This chapter presents the following topics:

VMWARE ESXi ........................................................................................................ 11

Round Robin - New Policy Recommendation ........................................................... 11

Changing Policy from Fixed Path to Round Robin ................................................... 12

10 EMC VPLEX Host Multipathing

Round Robin - Multipath Policy Recommendation

EMC VPLEX Host Multipathing 11

Changing Policy from Fixed Path to Round Robin

1. Open the vSphere CLI (recommended) or the service console.

– For ESXi 5.x:

– For ESXi 5.x:

12 EMC VPLEX Host Multipathing

For related information, refer to the following VMware Knowledgebase article:

For information on VPLEX host cross-cluster connect configuration using PowerPath

Specific recommendations and settings will be discussed in the following chapters.

EMC VPLEX Host Multipathing 13

This chapter presents the following topics:

Supported Versions and Settings ........................................................................... 16

14 EMC VPLEX Host Multipathing

PowerPath supports multiple paths to a logical device, enabling PowerPath to

 Automatic failover in the event of a hardware failure. PowerPath automatically

 Read, write MB/seconds per LUN

EMC VPLEX Host Multipathing 15

 Expanded support for standby: autostandby—An autostandby feature has been

Supported Versions and Settings

Before configuring VPLEX, complete the following on each host:

 Confirm that all necessary remediation has been completed.

PowerPath is highly recommended for the host cross-cluster connect configuration.

The load-balancing and failover policy apply to:

16 EMC VPLEX Host Multipathing

Chapter 4 Native MPIO

This chapter presents the following topics:

Microsoft Windows ................................................................................................ 31

EMC VPLEX Host Multipathing 17

18 EMC VPLEX Host Multipathing

 If a host is running EMC PowerPath, confirm that the load-balancing and